This application has three specific aims the first of which describes a systematic approach for testing the correlation between UVD (unintegrated viral DNA) and clinical status. Frozen peripheral blood monocytic cells (PBMC) (1,500 samples) representing more than 700 well-characterized patients will be tested for UVD levels. These data will be correlated with clinical and laboratory data available for this well-characterized patient population. Rigorous statistical analysis of the data will establish the validity of UVD as a marker for disease progression in AIDS. Analysis of UVD could also provide fresh insight into the mode of antiviral action in vivo. Therapeutic agents designed to reduce HIV replication and infection should also reduce the level of UVD. Because UVD is transient in the newly-infected cell, data on drug action could be obtained over a short testing period as compared to the longer intervals required to observe changes in surrogate markers of infection. The second goal of the investigator encompasses a plan to assess drug efficacy based on changes in UVD. This will be done by temporal analyses of UVD-positive patients receiving drug for the first time and on patients who are discontinuing drug therapy. In addition to its role in acute infection of cells, unintegrated DNA is also a feature in latent macrophage infection, the subject of the last specific aim. The structural features of UVD in latently-infected monocytic cells will be examined by molecular cloning, DNA sequencing and transfection studies. In order to investigate the role of these UVD species in HIV infection, the structure of viral DNA in central nervous system tissue will be examined. In this tissue, the principal infected cell type is the macrophage and the presence of UVD has already been documented.